U.S. patent number 6,845,326 [Application Number 09/706,747] was granted by the patent office on 2005-01-18 for optical sensor for analyzing a stream of an agricultural product to determine its constituents.
This patent grant is currently assigned to NDSU Research Foundation. Invention is credited to Suranjan Panigrahi, Guangjun Zhang.
United States Patent |
6,845,326 |
Panigrahi , et al. |
January 18, 2005 |
Optical sensor for analyzing a stream of an agricultural product to
determine its constituents
Abstract
An optical sensor for use in measuring constituents of an
agricultural product. An optical sensing window passes a stream of
the agricultural product, and a radiation source irradiates the
stream as it passes through the optical sensing window. A receiver
receives radiation transmitted through the stream and converts it
into a corresponding electrical signal using a spectrometer. The
electrical signal is digitized to produce a series of data points
corresponding to particular wavelengths. A processor normalizes the
data points using a reference value in order to generate processed
data points that can be used to predict a constituent content of
the agricultural product.
Inventors: |
Panigrahi; Suranjan (Fargo,
ND), Zhang; Guangjun (Beijing, CN) |
Assignee: |
NDSU Research Foundation
(Fargo, ND)
|
Family
ID: |
33568468 |
Appl.
No.: |
09/706,747 |
Filed: |
November 7, 2000 |
Current U.S.
Class: |
702/22;
250/339.02; 250/339.11; 250/339.12; 356/141.3; 356/300; 356/73.1;
702/134; 702/33; 702/40; 702/49 |
Current CPC
Class: |
G01J
3/02 (20130101); G01J 3/0291 (20130101); G01J
3/28 (20130101); A01D 41/1277 (20130101); G01N
21/85 (20130101); A01B 79/005 (20130101); G01N
21/31 (20130101); G01N 33/02 (20130101) |
Current International
Class: |
G01N
21/31 (20060101); G01J 3/02 (20060101); G01J
3/00 (20060101); G01J 3/28 (20060101); G01N
33/02 (20060101); G06F 019/00 () |
Field of
Search: |
;702/22,33,40,49,134
;250/339.02,339.12,341.8,339.11,360.1
;356/141.3-141.4,139.04-139.08,300,326,73.1 |
References Cited
[Referenced By]
U.S. Patent Documents
Foreign Patent Documents
|
|
|
|
|
|
|
0 388 082 |
|
Sep 1990 |
|
EP |
|
WO 99/40419 |
|
Aug 1999 |
|
WO |
|
Other References
Randy Nouis, Predicting the Ninety-Fifth Percentile Dust
Environment for Passenger Vehicles in the Continental United
States, Mar. 1-5, 1993, pp. 1-11. .
C. S. Chang et al., Grain Flow Regulator for Dust Emission Control,
Nov.-Dec., 1985, pp. 2059-2062. .
Joseph A. Borgia et al., Pressure Drop and Flow Characteristics for
a Heavy-Duty Air Filter During Dust Loading, Nov. 16-19, 1987, pp.
1-15. .
Charles O. Reinhart et al., Measurement of Engine Air Cleaner
Efficiency Using Airborne Particle Size Analysis, Sep. 12-15, 1983,
pp. 1-8. .
Philip C. Williams, et al.; Determination of Protein and Moisture
in Wheat and Barley by Near-Infrared Transmission, 1985, pp.
239-244. .
K. H. Norris et al., Optimization of Mathematical Treatments of Raw
Near-Infrared Signal in the Measurement of Protein in Hard Red
Spring Wheat. I. Influence of Particle Size, 1984, pp. 158-165.
.
Renfu Lu et al., Determination of Firmness and Sugar Content of
Apples Using Near-Infrared Diffuse Reflectance, 24 pages. .
F. De Lene Mirouze et al., Quantitative Analysis of Glucose Syrups
by ATR/FT-IR Spectroscopy, 1993, pp. 1187-1191. .
Veronique Bellon-Maurel et al., Quantitative Analysis of Individual
Sugars during Starch Hydrolysis by FT-IR/ATR Spectrometry. Part I:
Multivariate Calibration Study--Repeatibility and Reproducibility,
1995, pp 556-562. .
Suranjan Panigrahi, et al., On-The-Go Sensing Techniques for Sugar
Determination of Sugarbeet in the Field, pp 176-178. .
E. K. Kemsley et al., Quantitative analysis of sugar solutions
using infrared spectroscopy, 1992, pp299-304. .
Dr. Daniel S. Humburg et al., Spectral Analysis of Sugar Beet
Canopy for Spatial and Temporal Quantification of Sugar Content,
Quality, and Disease, Dec. 14, 1999, pp. 1-6. .
Dave Berard, Protein Monitor on the Market, 07/97, pp. 18-19. .
Carol R. Dumas, New Milestone Monitor Analyzes Grain Quality On The
Go, Feb. 14, 1997, pp. 1-4. .
Jerry Workman, Jr., A Compact Reference for Practitioners, pp.
Contents and 423-435. .
Food Processing Automation, May 6, 1990, pp. Table of Contents and
103-114. .
Donald A. Burns, Handbook of Near-Infrared Analysis, pp. Contents
53-106. .
P. C. Williams et al., Effect of Mutual Interactions on the
Estimation of Protein and Moisture in Wheat, Nov. 24, 1982. .
Russell Tkachuk, Ph.D., Protein Analysis of Whole Wheat Kernels by
Near Infrared Reflectance .sup.1,2. .
G. Downey et al., Protein Testing of Wheat by Near Infrared
Reflectance. .
Frederic Cadet et al., Direct Spectroscopic Sucrose Determination
of Raw Sugar Cane Juices, 1997, pp 166-171. .
Gianluigi Marchetti, Application of a NIR on-line automatic
analyzer system in a beet sugar factory, 1990, pp 210-215. .
Nils Berding et al., Crop Ecology, Production & Management,
1991, pp 1017-1023. .
Nils Berding et al.; Near Infrared Reflectance Spectroscopy for
Analysis of Sugarcane from Clonal Evaluation Trials: II. Expressed
Juice, 1991, pp 1024-1028. .
Roberto Giangiacomo et al., Near Infrared Spectrophotometric
Determination of Individual Sugars in Aqueous Mixtures, 1986, pp
679-683. .
Stephen R. Delwiche et al., Classification of Hard Red Wheat by
Near-Infrared Diffuse Reflectance Spectroscopy, 1993, pp. 29-35.
.
Huaipu Song et al., Neural Network Classification of Wheat Using
Single Kernel Near-Infrared Transmittance Spectra, Oct. 1995, pp.
2927-2934. .
Stephen R. Delwiche, Single Wheat Kernel Analysis by Near-Infrared
Transmittance: Protein Content, 1995, pp. 11-16. .
Philip C. Williams, Application of Near Infrared Reflectance
Spectroscopy to Analysis of Cereal Grains and Oilseeds, 1975, pp.
561-576. .
Near-Infrared Reflectance Method for Protein Determination, pp.
1-2. .
Near-Infrared Method for Protein Content in Whole-Grain Wheat, pp.
1-3. .
Innovative Protein Monitoring, pp. 1-2. .
File History--U.S. Appl. No. 5,751,421. .
J. Sorvaniemi et al., Using Partial Least Squares Regression and
Multiplicative Scatter Correction of FT-NIR Data Evaluation of
Wheat Flours, 1993, pp 251-258. .
Suming Chen et al., "Neural Network Analysis of Sugar Content in
Fruit Juice", Jul. 18-21, 1999, pp. Title page thru 12. .
Renfu Lu et al., "Determination of Sugar Content and Firmness of
Apples Using Near-Infrared Diffuse Reflectance", 07/9-12, 00, pp.
Title page thru 16. .
K. J. Kaffka et al., "Attempts to Determine Oil, Protein, Water and
Fiber Content in Sunflower Seeds by the NIR Technique", 1983, pp
117-129. .
Essex E. Finney, Jr. et al., "Determination of Moisture in Corn
Kernals by Near-Infrared Transmittance Measurements", 1978, pp
581-584. .
Wang-Sheng Li et al., "Determination of Rough Rice Quality by a
Portable Near-Infrared Spectroscopy", 1997, (5) pages. .
D. T. Lamb et al., "Moisture Determination in Single Soybean Seeds
by Near-Infrared Transmittance", 1991, pp 2123-2129. .
P. C. Williams et al., "Influence of Temperature on Estimation of
Protein and Moisture in Wheat by Near-Infrared Reflectance", 1982,
pp473-477. .
D. Wang et al., "Effect of Wheat Kernel Size and Orientation on
Reflectance Spectra and Single Kernel Color Classification", 1997,
pp 1-34. .
S. R. Delwiche, "Measurement of Single-Kernel Wheat Hardness Using
Near-Infrared Transmittance", 1993, pp1431-1437. .
M. R. Campbell et al., "Whole Grain Amylose Analysis in Maize Using
Near-Infrared Transmittance Spectroscopy", 1997, pp300-303. .
Stermer et al., "Infrared Reflectance Spectroscopy for Estimation
of Moisture of Whole Grain", 1997, pp345-351. .
F. E. Dowell et al., "Automated Single Wheat Kernel Quality
Measurement Using Near-Infrared Reflectance", 1997, pp 1,3,5,7 and
9. .
B. G. Osborne, "Recent Progress in the Application of NIR to the
Measurement of Quality Parameters in Flour", 1982, pp577-581. .
J. S. Shenk, "How NIR Can Help in Measuring Forage Quality for
Breeding and Utilization Programs", (3) pages. .
Gerard Downey, "Estimation of Moisture in Undried Wheat and Barley
by Near Reflectance", 1985, pp951-958. .
Doninique Bertrand et al., "Application of Principal Component
Analysis to the Prediction of Lucerne Forage Protein Content and in
vitro Dry Matter Digestibility by NIR Spectroscopy", 1987,
pp299-307. .
Shuso Kawamura et al., "Determining Undried Rough Rice Constituent
Content Using Near-Infrared Transmission Spectroscopy", 1997, pp
cover, 1,3,5 and 7. .
Holger M. Jaenisch et al., "Instrumentation to Measure the Near-IR
Spectrum of Small Fruits", 1990, pp 162-166. .
Paul Geladi et al., "Partial Least-Squares Regression: A Tutoria",
1985, pp 1-17. .
Constantinos Goutis, "Second-Derivative Functional Regression with
Applications to Near Infra-Red Spectroscopy", 1998, pp 103-114.
.
Milling Feed and Fertiliser, "FMBRA takes the Grind out of
Measuring Moisture", pp 34-35. .
John S. Shenk et al., "Near Infrared Reflectance Analysis with
Single-and Multiproduct Calibrations", 1993, pp582-584. .
D. T. Williams et al., "The Derivative Spectrometer", 1970, pp
1597-1605. .
F. E. Barton et al., "The Calibration of NIR Reflectance
Spectrometer for the Determination of Diverse Compositional
Parameters", 1988, pp 768-773. .
R. J. Barnes et al.; "Standard Normal Variate Transformation and
De-Trending of Near-Infrared Diffuse Reflectance Spectra", 1989,
pp772-777. .
Kurt C. Lawrence et al., "Sensing Wheat Moisture Content
Independent of Density", 1997, pp Cover Page, 1,3,5,7,9,11,13 and
15. .
Abraham Savitzky et al., "Smoothing and Differentiation of Data by
Simplified Least Squares Procedures", 1964, pp1627-1639. .
Principle Introduction to PROTRONICS, (8) pages. .
"Comments on the Savitzky-Golay Convolution Method for
Least-Squares Fit Smoothing and Differentiation of Digital Data",
1978, pp1383-1386. .
Gerald S. Birth et al., "Interaction Between Light and Natural
Materials: Laboratory Demonstrations", (4) Cover Pages, and pp1-36.
.
Gerald S. Birth, "How Light Interacts with Foods", (6) pages. .
H. Martens et al., "Partial Least Squares Regression: A New
Two-Stage NIR Calibration Method", 1982, pp607-647. .
Tormod N.ae butted.s et al., "Comparison of Multivariate
Calibration and Discriminant Analysis in Evaluating NIR
Spectroscopy for Determination of Meat Tenderness", 1996,
pp350-357. .
G. Asimopoulos et al., "On-Line Monitoring of Dairy Products with
the Use of NIR Technology", pp 266-271. .
"Comments on Smoothing and Differentiation of Data by Simplified
Least Square Procedure", 1972, pp 1906-1910. .
Edward K. Baldwin, Ph.D., "Calibrating Near Infrared Instruments
for On-Line Food Processing Measurements", pp 252-265. .
P. D. Wilson et al., "Polynomial Filters of any Degree", 1981, pp
599-603. .
L. P. McDermott, "The Benefits and Pitfalls of Applying Near
Infrared Analysis On-Line", pp 103-114. .
B. G. Osborne, "Monitoring the accuracy of NIR Instruments", 1987,
pp 515-521. .
James R. Long, "Spectroscopic Calibration and Quantitation Using
Artificial Neural Networks", 1990, pp 1791-1797. .
Phil Williams et al., "Near-Infrared Technology in the Agricultural
and Food Industries", (18) pages. .
B. G. Osborne et al., "Practical NIR Spectroscopy with Application
in Food and Beverage Analysis", (16) pages. .
Jerry Workman, Jr. et al., "Applied Spectroscopy A Compact
Reference for Practitioners", (10) pages. .
Reginald H. Wilson, "Spectroscopic Techniques for Food Analysis",
(9) pages. .
Kerstin Jakobsson, Foss Tecator AB, "Reliable Grain Trading Based
on Professional Decisions", 1998 (4) pages..
|
Primary Examiner: Barlow; John
Assistant Examiner: Vo; Hien
Attorney, Agent or Firm: Snell & Wilmer LLP
Parent Case Text
REFERENCE TO RELATED APPLICATIONS
The present application is a continuation-in-part of U.S.
provisional patent application Ser. No. 60/164,161, filed Nov. 8,
1999, and entitled "Optical Analysis of Grain Stream," which is
incorporated herein by reference.
The present application is related to the following application,
which is incorporated herein by reference: U.S. provisional patent
application Ser. No. 60/175,636, filed Jan. 12, 2000, and entitled
"On-The-Go Sugar Sensor for Determining Sugar Content During
Harvesting."
Claims
What is claimed is:
1. An apparatus for measuring a constituent content of an
agricultural product, comprising: a device for forming a stream of
the agricultural product; an optical sensing window in the device
allowing passage of radiation from a radiation source to pass
through the stream of agricultural product to a receiver, the
optical sensing window being configured, in comparison to the
device, to provide a narrower passageway for the stream of
agricultural product to provide for a more uniform consistency in
the stream of agricultural product; a radiation source for
irradiating the stream of the agricultural product as the stream of
the agricultural product passes through the optical sensing window;
a receiver for receiving radiation transmitted from the radiation
source through the stream of agricultural product and the optical
sensing window, and for converting the received radiation into a
corresponding electrical signal; and a computer, coupled to the
receiver, for receiving the electrical signal and for processing
the electrical signal to generate data for use in determining a
constituent content of the agricultural product.
2. The apparatus of claim 1 wherein the receiver includes: a fiber
optic cable; a sensor head for receiving the radiation transmitted
from the radiation source through the stream of agricultural
product and the optical sensing window, and for focusing the
received radiation onto the fiber optic cable; and a spectrometer,
coupled to the fiber optic cable, for converting the received
radiation into the corresponding electrical signal.
3. The apparatus of claim 2 wherein the fiber optic cable comprises
a single fiber optic cable.
4. The apparatus of claim 2 wherein the sensor bead includes: a
fiber optic probe coupled to the fiber optic cable; and a plurality
of optical lenses positioned between the optical sensing window and
the fiber optic cable for focusing the received radiation onto the
fiber optic probe.
5. The apparatus of claim 1, further including a housing for
containing the device, the radiation source, the optical sensing
window, and the receiver.
6. The apparatus of claim 5, further including a fan mounted within
the housing.
7. The apparatus of claim 1, further including an inlet, coupled to
the device, for attachment to a source providing the agricultural
product.
8. The apparatus of claim 7 wherein the inlet is configured to
receive the agricultural product from a combine.
9. The apparatus of claim 1 wherein the device receives the
agricultural product from a grin pipe.
10. The apparatus of claim 1 wherein the optical sensing window
includes: an inner wall; an outer wall; and a pair of side walls,
wherein the inner wall, the outer wall, and the pair of side walls
are joined to form a passageway for the agricultural product.
11. The apparatus of claim 10 wherein: the inner wall is formed
from an optically transparent material; and the outer wall and the
pair of side walls are formed from an opaque material.
12. The apparatus of claim 11 wherein the inner wall includes: a
front planar section; and a pair of curved edges.
13. The apparatus of claim 10 wherein the outer wall includes a
transparent aperture for permitting the radiation to pass through
the stream of the agricultural product and to the receiver.
14. A method for measuring constituent contents of an agricultural
product, comprising: forming a stream of the agricultural product;
passing the stream of the agricultural product through an optical
sensing window allowing passage of radiation from a radiation
source to pass through the stream of agricultural product to a
receiver, the optical sensing widow being configured, in comparison
to the device, to provide a narrower passageway for the stream of
agricultural product to provide for a more uniform consistency in
the stream of agricultural product; irradiating the stream of the
agricultural product as the stream of the agricultural product
passes through the optical sensing window; receiving radiation
transmitted from the radiation source through the stream of
agricultural product and the optical sensing window, and converting
the received radiation into a corresponding electrical signal; and
receiving the electrical signal and processing the electrical
signal to generate data for use in determining a constituent
content of the agricultural product.
15. The method of claim 14 wherein the receiving radiation step
includes: receiving the radiation transmitted through the stream of
the agricultural product and focusing the received radiation onto a
fiber optic cable; and converting the received radiation from the
fiber optic cable into the corresponding electrical signal.
16. The method of claim 15 wherein the receiving the radiation step
includes focusing the received radiation onto a single fiber optic
cable.
17. The method of claim 14 wherein the receiving radiation step
includes using a plurality of optical lenses for focusing the
received radiation onto a fiber optic probe coupled to the fiber
optic cable.
18. The method of claim 14, further including providing an inlet
for attachment to a source providing the agricultural product.
19. The method of claim 18 wherein the providing step includes
configuring the inlet to receive the agricultural product from a
combine.
20. The method of claim 13, further including receiving the
agricultural product from a grain pipe.
21. A method for converting a light signal into an electrical
signal for use in predicting a constituent content of an
agricultural product, comprising: receiving a light signal from an
agricultural product; converting the light signal into an
electrical signal; digitizing the electrical signal to produce a
plurality of data points; and normalizing the data points using a
reference signal value to produce a plurality of normalized data
points, the normalized data points having values related to a
constituent content of the agricultural product, wherein the
reference signal value is related to a magnitude at a wavelength of
the light signal substantially unaffected by the constituent
content.
22. The method of claim 21 wherein the normalizing step includes
using, as the reference signal value, a value derived from the
magnitude using a mathematical function.
23. The method of claim 21 wherein the normalizing step includes
using as the reference signal value a value related to magnitudes
at a plurality of wavelengths including the wavelength of the light
signal substantially unaffected by the constituent content.
24. The method of claim 21 wherein the normalizing step includes
using as the reference signal value an average magnitude value of a
range of magnitude values at a pair of wavelengths centered around
the reference wavelength.
25. The method of claim 21, further including predicting protein
content of the agricultural product using the plurality of
normalized data points.
26. The method of claim 21, further including: receiving
geographical coordinates corresponding with a geographical location
of the agricultural product; and associating the geographical
coordinates with the constituent content of the agricultural
product.
27. The method of claim 26, further including: receiving a
plurality of geographical coordinates corresponding with
geographical locations of a plurality of agricultural products for
which the constituent contents are predicted using the method; and
associating the plurality of geographical coordinates with the
constituent contents of the plurality of agricultural products.
28. The method of claim 27, further including generating a map of
the constituent content of the agricultural products using the
plurality of geographical coordinates and the associated
constituent contents.
29. The method of claim 28 wherein the generating step includes
generating a grid map.
30. The method of claim 28 wherein the generating step includes
generating a contour map.
31. The method of claim 21, further including calculating average
values for the data points and wherein the normalizing stop
includes normalizing the average values.
32. The method of claim 21, further including linearizing the
normalized data points.
33. The method of claim 32 wherein the linearizing step includes
calculating a logarithm of each of the data points.
34. The method of claim 21 wherein the receiving step includes
receiving the light signal from moving stream of the agricultural
product.
35. The method of claim 21 wherein the receiving step includes
receiving the light signal from a stopped stream of the
agricultural product.
36. A method for converting a light signal into an electrical
signal for use in predicting a constituent content of an
agricultural product, comprising: receiving a light signal from an
agricultural product; converting the light signal into an
electrical signal; digitizing the electrical signal to produce a
plurality of data points; and normalizing the data points using a
reference signal value to produce a plurality of normalized data
points, the normalized data points having values related to a
constituent content of the agricultural product, wherein the
reference signal value corresponds with a magnitude of a received
light signal at a specific wavelength without being transmitted
through the agricultural product.
37. A method for converting a light signal into an electrical
signal for use in predicting a constituent content of an
agricultural product, comprising: receiving a light signal from an
agricultural product; converting the light signal into an
electrical signal; digitizing the electrical signal to produce a
plurality of data points; and normalizing the data points using a
reference signal value to produce a plurality of normalized data
points, the normalized data points having values related to a
constituent content of the agricultural product, wherein the
reference signal value corresponds with a magnitude of a received
light signal transmitted through a gating mechanism and without
being transmitted through the agricultural product.
38. An apparatus for measuring a constituent content of an
agricultural product, comprising: a device for forming a stream of
the agricultural product; an optical sensing window in the device
for passing the stream of the agricultural product; a radiation
source contained within the housing for irradiating the stream of
the agricultural product as the stream of the agricultural product
passes through the optical sensing window; a receiver for receiving
radiation transmitted through the stream of the agricultural
product and for converting the received radiation into a
corresponding electrical signal; and a computer, coupled to the
receiver, for receiving the electrical signal and for processing
the electrical signal to generate data for use in determining a
constituent content of the agricultural product, the computer
operating to: digitize the electrical signal to produce a plurality
of data points; and normalize the data points using a reference
signal value to produce a plurality of normalized data points, the
normalized data points having values related to a constituent
content of the agricultural product, wherein the reference signal
value is related to a magnitude at a wavelength of the light signal
substantially unaffected by the constituent content.
39. The apparatus of claim 38 wherein the computer operates to use,
as the reference signal value, a value derived from the magnitude
using a mathematical function.
40. The apparatus of claim 38 wherein the computer operates to use
as the reference signal value a value related to magnitudes at a
plurality of wavelengths including the wavelength of the light
signal substantially unaffected by the constituent content.
41. The apparatus of claim 38 wherein the computer operates to use
as the reference signal value an average magnitude value of a range
of magnitude values at a pair of wavelengths centered around the
reference wavelength.
42. The apparatus of claim 38 wherein the computer operates to
predict protein content of the agricultural product using the
plurality of normalized data points.
43. The apparatus of claim 38 wherein the computer operates to
normalize the data points selectively using one of the following
plurality of reference signal values: a magnitude at a wavelength
of the light signal substantially unaffected by the constituent
content, a magnitude of a received light signal without being
transmitted through the agricultural product at specific
wavelengths, or a magnitude of a received light signal transmitted
through a gating mechanism and without being transmitted through
the agricultural product.
Description
FIELD OF THE INVENTION
The present invention relates to a method and apparatus for
optically analyzing a stream of an agricultural product in order to
determine constituents of the product.
BACKGROUND OF THE INVENTION
Systems are known in the art for the optical analysis of a stream
of grain. As the grain is harvested in the field, a light source
passes light through the grain stream. The transmitted light is
detected by a receiver and processed by a computer under software
control. By comparing the spectral absorption with values
representing known absorption, the grain can be analyzed to
determine its constituents. Examples of systems for handling grain
are disclosed in U.S. Pat. Nos. 5,343,761 and 5,369,603, both of
which are incorporated herein by reference.
To make that determination, a reference value is used. For example,
the light can be transmitted to a white ceramic tile without grain
present. The light reflected and received from the tile provides a
reference value for use in analyzing the light as reflected from
the grain. The use of a tile can have disadvantages. It can become
dusty, particularly from the grain, and the dust can affect the
light reflected from the tile. Therefore, the dust can alter the
reference value and affect the accuracy of the analysis. In
addition, the use of a tile to obtain a reference value only works
for analyzing a signal reflected from the grain, since the
reference is obtained from a reflected signal. Other references
must be used for analyzing a signal transmitted through the
grain.
Accordingly, a need exists for improvements in these systems for
optical analysis of agricultural products.
SUMMARY OF THE INVENTION
An apparatus consistent with the present invention is used for
measuring a constituent content of an agricultural product. It
includes a device for forming a stream of the agricultural product,
and an optical sensing window for passing the stream of the
agricultural product. A radiation source irradiates the stream of
the agricultural product as it passes through the optical sensing
window, and a receiver receives radiation transmitted through the
stream and converts it into a corresponding electrical signal. A
computer receives the electrical signal and processes it in order
to generate data for use in determining a constituent content of
the agricultural product.
A method consistent with the present invention converts a light
signal into an electrical signal for use in predicting a
constituent content of an agricultural product. It includes
receiving a light signal from an agricultural product and
converting the light signal into an electrical signal. The
electrical signal is digitized to produce a plurality of data
points, which are then normalized using a reference signal value to
produce a plurality of normalized data points. The reference signal
value corresponds with a magnitude at a wavelength of the light
signal substantially unaffected by the constituent content. The
normalized data points have information related to a constituent
content of the agricultural product and can thus be used to predict
the constituent content.
BRIEF DESCRIPTION OF THE DRAWINGS
The accompanying drawings are incorporated in and constitute a part
of this specification and, together with the description, explain
the advantages and principles of the invention. In the
drawings,
FIG. 1 is a diagram illustrating optical analysis of a stream of
agricultural products consistent with the present invention;
FIGS. 2A and 2B are side and rear views of a conventional combine,
illustrating use of an optical sensor for analysis of a stream of
an agricultural product on the combine during harvesting of the
product;
FIG. 3 is a block diagram of a system including an optical sensor
for analyzing a stream of an agricultural product;
FIG. 4 is a diagram of a side cover in the system of FIG. 3;
FIG. 5 is a perspective diagram of a detector box in the system of
FIG. 3;
FIG. 6 is a perspective view of an optical sensing window in the
system of FIG. 3;
FIG. 7 is a side view of the optical sensing window;
FIG. 8 is a top view of the optical sensing window;
FIG. 9 is a front view of the optical sensing window;
FIG. 10 is a diagram of exemplary components of a computer for use
with the optical sensor;
FIG. 11 is a flow chart of a main routine for processing of a
signal from the optical sensor;
FIG. 12 is a diagram of a main screen for use with the optical
sensor;
FIG. 13 is a flow chart of a determine constituent routine for use
with the optical sensor,
FIG. 14 is a flow chart of an analyze sampled agricultural product
routine;
FIG. 15 is a graph illustrating spectral components of a reflected
signal from the optical sensor and a reference wavelength;
FIG. 16 is a graph illustrating spectral components of the
reflected signal and a band of reference wavelengths;
FIG. 17 is a flow chart of a method for analyzing a sampled
agricultural product according to an alternate embodiment;
FIG. 18 is a screen for displaying real-time constituent data;
FIG. 19 is a flow chart of a generate map routine for providing a
visual indication of protein content within a field of an
agricultural product;
FIG. 20 is an example of a grid map illustrating protein content
throughout a field of an agricultural product; and
FIG. 21 is an example of a contour map illustrating protein content
throughout a field of an agricultural product.
DETAILED DESCRIPTION
Overview
FIG. 1 illustrates a system 10 for optical analysis of a stream of
an agricultural product consistent with the present invention. In
system 10, a stream of an agricultural product 14 is produced
within a conduit 12. From the stream of agricultural product 14 a
portion of it 16 is directed to an optical sensor 18. Following
analysis of portion 16, it is returned as portion 20 to the stream
of agricultural product 14 within conduit 12. Accordingly, optical
sensor 18 analyzes a portion of a stream of an agricultural product
from a conduit of the product and then returns that portion, or
most of it, back to the stream of the agricultural product.
The stream of agricultural product 14 and conduit 12 can be derived
from a variety of sources. For example, conduit 12 may represent a
pipe within a combine as it harvests a field of an agricultural
product such as grain. In addition, conduit 12 may alternatively
represent a pipe transporting grain within a food processing
facility. Therefore, optical sensor 18 provides an advantage of
operating on a stream of grain at various locations, such as in the
field or within an assembly line of a food processing plant, and
provides in real-time or near real-time an indication of a
constituent content of the stream of agricultural product.
Optical system 18 analyzes portion 16 of the agricultural product
using optical sensing components. As further explained below,
optical sensor 18 transmits light through the stream of the
agricultural product and receives light from the agricultural
product, either transmitted through or reflected from it. The
received light signal is converted to an electrical signal and
transmitted on line 24 to a computer 26. The term "light signal" is
intended to include visible light signals, invisible light signals,
or both. Also, the term "received light signal" includes a light
signal reflected from or transmitted through an agricultural
product. Computer 26 digitizes the electrical signal to produce
data points, and processes the data points to provide an indication
of a constituent content based upon the signal and upon how various
constituents affect the light absorption properties of the
agricultural product. Once the constituent content is predicted,
computer 26 may store and output an indication of that constituent
content in a variety of forms. During the operation, because of
random size configurations of grain samples and the intra-granular
space, the transmitted light through the product could saturate the
receiver. In that case, computer 26 can be programmed to detect
this event and change the data acquisition parameter, such as
integration time, of the spectrometer to acquire the raw spectral
signal. Alternatively, on detecting this event, computer 26 can be
programmed to trigger a motor 48 to advance the grain or
agricultural product in the sensing window for a very small amount
of time. This will change the configuration of the sample (product)
in the sensing window, and the new configuration might not saturate
the receiver.
Optical sensor 18 can be used to analyze various types of
agricultural products, such as grain or wheat. Also, optical sensor
18 along with computer 26 can provide for prediction of a variety
of constituents such as protein, starch, fiber, or moisture
content. Also, multiple constituents can be determined through
analysis of the same signal from optical sensor 18; this feature
eliminates, for example, the need for separate sensors, although
different sensors can still be used if desired.
FIG. 2A illustrates use of optical sensor 18 attached to a
conventional combine 30 for determining constituent contents of an
agricultural product during harvesting. Conventional combine 30
includes a cab 31 for a driver to operate the combine. A combine
header attachment 32 is used for attaching various devices to
harvest the agricultural product. A clean grain elevator 33 can
generate a stream of the agricultural product, and optical sensor
18 can thus remove a portion (16) of the stream, sample it, and
return it (20) to clean grain elevator 33. A grain tank 35 holds
the harvested agricultural product, and the product can be
discharged through unloading auger 34 via an auger 36. As
illustrated, optical sensor 18 in this example can be attached to a
side of the combine (FIG. 2A) or a rear of the combine (FIG.
2B).
Optical Sensor System
A system to predict the protein or other constituent contents of
agricultural products is illustrated with reference to FIGS. 3-9.
As shown in FIG. 3, an inlet 41 of a material handling system
typically includes particular pipes and accessories to connect with
a source of an agricultural product depending upon a particular
application and the requirements necessary to branch off a stream
of the agricultural product for analysis. For example, inlet 41 can
be attached to an auger, a clean grain elevator, or any outlet of a
storage bin, and the term "inlet" includes any mechanism for
assisting in forming a stream of an agricultural product for
analysis. The grain or agricultural product entering through inlet
41 moves through a grain passage 43, bounded by an inner
transparent wall 44 and a metallic outer wall 45. The grain
entering through inlet 41 passes through an optical sensing window
46, as illustrated by arrow 130.
A proximity sensor 42 is connected to a control unit 47, which is
also connected to an electric motor 48 operated by direct current
(DC) power source 49. Electric motor 48 is mounted within an
enclosure box of an auger 50 containing discharger auger 51, which
is driven by motor 48. Auger 51 through an outlet 52 can discharge
the grain from the system back to the original stream of grain or
any other user-defined location.
An illumination chamber 53 is bounded by transparent wall 44 and a
vertical opaque wall 57. A base 54 is mounted on wall 57. An
illumination source 56, implemented with a lamp or other device
providing a light signal, is attached by a lamp holder 55 and is
connected to the power source and control box 59 through a cable
88. Lamp 56 may be implemented with, for example, a
tungsten-halogen lamp.
A sensor body 61 includes air inlet passages 83. Sensor body 61 is
attached to a DC fan 60. A side cover 62 of sensor body 61 (shown
in FIG. 4) has another fan 63 mounted on the cover and is operated
by a DC power source 64.
A sensor head 72 is composed of optical passage 65, optically
isolated from outer environment by metallic covers 73 and a
detector box 66. Sensor head 72 is attached to sensor body 61 by a
mount 71. The tip of a fiber optic probe 67 is mounted on detector
wall 74 (see FIG. 5) of detector box 66. Fiber optic cable 68 is
connected to a portable spectrometer 69 including a diffraction
grating and an array of charged coupled device (CCD) detectors.
Spectrometer 69 is coupled to with a computer 70. Detector box 66
and spectrometer 69 together form a receiver for converting the
light signal or received radiation into a corresponding electrical
signal.
As shown in FIG. 5, detector box 66 is composed of a front lens
wall 75, a detector wall 74, a base plate 77, and a top cover 76.
Front lens wall 75 contains two lenses, 81 and 82 (in series)
arranged between three retainers 78, 79, and 80. Cover 76 includes
a top 93, and sides 94 and 95. In use, cover 76 may be fastened to
base plate 77 by inserting fasteners through apertures on sides 94
and 95 such as aperture 120. The fasteners may then engage
corresponding apertures on the bottom of base plate 77, such as
aperture 121, for receiving a fastener through aperture 120. As
shown, the bottom portions of sides 94 and 95 include a plurality
of such apertures for receiving a plurality of fasteners for
attachment to base plate 77.
Also, base plate 77 includes a plurality of apertures such as
apertures 89, 90 and 91 for receiving fasteners and attaching base
plate 77 to a bracket 71 (see FIG. 3). Although bracket 71 is shown
as angled in the side view of FIG. 3, it can also be implemented
with a perpendicular bracket or other configurations of brackets to
support detector box 66 in this embodiment. When cover 76 is
mounted on base plate 77, top 93 is set flush against top portions
96 and 98, and sides 94 and 95 are set flush against sides portion
74, 75, 97, and 99. Therefore, cover 76 provides for blocking of
ambient light within detector box 66.
Detector wall 74 includes an aperture 92 for receiving and mounting
an end of fiber optic cable at point 67. In addition, front lens
wall 75 includes an aperture 132 for receiving the lens for
focusing the received light, as illustrated by line 133, on an end
of the fiber optic cable in aperture 92.
Optical Sensing Window
FIG. 6 is a perspective view providing more detail of optical
sensing window 46 in the system of FIG. 3. FIGS. 7-9 are,
respectively, side, top, and front views also providing more detail
of optical sensing window 46. The dimensions provided in FIGS. 7-9
arc in inches and illustrate a preferred dimensional configuration
for optical sensing window 46. An optical sensing window includes
any device for providing a way to pass a light signal through a
stream of an agricultural product, and optical sensing window 46 is
one such example. Also, optical sensing windows can be configured
according to different dimensions than those illustrated.
As shown in FIGS. 6-9, optical sensing window 46 is formed by inner
transparent wall 44, and outer sensing wall 86, located behind
metallic outer wall 45. The two side walls 85 and 103 of the
optical sensing window are composed of opaque materials, and they
connect the inner and outer walls 44 and 86. Grain passage 43 is
empty space formed by the inner transparent wall 44, outer sensing
wall 86, and two side walls 85 and 103. A circular area 87 on the
metallic outer wall 45 defines the effective sensing region through
which the transmitted beam passes to the sensor head.
An inner transparent wall 44 includes a front planar section 100
and the angled sides 101 and 102. A plurality of fasteners 104,
105, 106, and 107 are used for fastening together a first portion
of the window. Another plurality of fasteners 108, 109, 110 and 111
are used for fastening together a second portion of the window.
The stream of agricultural product passes through the window as
shown by arrow 130. The structure of the optical sensing window
provides for a narrower passageway for the stream of agricultural
product compared with inlet 41. The narrower passageway provides
for a more uniform consistency in the stream of agricultural
product and provides for movement of the agricultural product that
helps to prevent accumulation of dust or agricultural product on
the sensing region formed by circular area 87. The narrower
passageway also provides for a suitable thickness of the stream of
agricultural product allowing it to produce an optimum transmission
of light through the product. These features provide for more
consistent and accurate readings of the light signal from the
agricultural product
Optical Sensor Operation
The grain or agricultural product enters through inlet 41, and it
fills up grain passage 43 and 84 and the empty space in the auger.
When the level of grain reaches the level at which position or
proximity sensor 42 is located, the sensor outputs a signal. A
switch 15 selectively connects proximity sensor 42 either to
control unit 47 via line 19 connected to line 13, or to
spectrometer 69 via line 19 connected to line 17. When proximity
sensor 42 is connected to control unit 47, it triggers, via the
signal detecting the level of grain at the sensor control, unit 47
to turn on motor 48 and run auger 51. The running of the auger
allows the grain to move through auger 51 and out from the system
through outlet 52. The location of position sensor 42 along with
features of wall 44, optical sensing window 46, grain passage 43
and 84, and auger 51, allow the grain to move at a constant rate
for sampling the stream of agricultural product. This feature may
also help to eliminate dust build-up on the inner wall of optical
sensing window 46. Dust build-up can adversely affect performance
of the sensor.
Alternatively, when proximity sensor 42 is connected to
spectrometer 69, the signal from proximity sensor 42 can be
transmitted from spectrometer 69 to computer 70 for use by the
computer in controlling motor 48 via line 21. The signal from
proximity sensor 42 indicates that the stream of agricultural
product has reached a level of the sensor and is thus present in
sensing window 46. Upon receiving that signal, computer 70 can turn
off motor 48 via line 21 in order to stop the flow of agricultural
product and sample the stopped flow. After sampling, computer 70
can signal control unit 47 to turn on motor 48 and restart the
stream of agricultural product.
Accordingly, the system can sample a moving stream of agricultural
product in a continuous mode or sample a stopped stream in a
stop-and-go mode. Instead of controlling the motor to stop the
stream, the system can alternatively use a gating mechanism in
grain passage 43 to momentarily stop the flow. The mode can be
determined through the position of switch 15. Switch 15 can be
implemented, for example, with a mechanical switch manually
controlled by a user, or with an electromechanical switch connected
to computer 70 and controlled by a command signal from computer
70.
For the sampling, illumination source 56 transmits a beam
containing visible and near infrared (NIR) spectrum through the
flowing or stopped agricultural product in the optical sensing
window 46. The transmitted light passes through optical passage 65
and subsequently passes through detector box 66. The front wall of
optical system 75 of detector box 66 converges the transmitted
light or radiation on the tip of fiber optic probe 67. The
transmitted light/radiation is conveyed through fiber optic cable
68 to spectrometer 69. Spectrometer 69 with the use of the computer
70, under software control, records the spectral signature of the
transmitted light or radiation between 700-1100 nanometers (nm) in
this example for determining protein content. Other spectral ranges
may be used depending on the agricultural product and constituent
to be analyzed.
Computer Hardware and Related Components
FIG. 10 depicts a data processing system 150 with a computer 151
illustrating exemplary hardware components of computer 70. Computer
151 may include a connection with a network 160 such as the
Internet or other type of network, and it can include a wireline or
wireless connection with the network. For example, if the optical
sensor is used in a food processing facility, the computer may
include a wireline connection with a network to transmit stored
constituent data. On the other hand, if the optical sensor is used
on a combine, the computer may include a wireless connection with a
network to transmit the constituent data.
Computer 151 typically includes a memory 152, a secondary storage
device 154, a processor 155, an input device 156, a global
positioning system (GPS) 157, a display device 153, and an output
device 158. A GPS is known in the art and provides approximate
longitude and latitude coordinates for its geographic location
based upon triangulation of signals received from GPS satellites.
Memory 152 may include random access memory (RAM) or similar types
of memory, and it may store one or more applications 159 for
execution by processor 155. Secondary storage device 154 may
include a hard disk drive, floppy disk drive, CD-ROM drive, or
other types of non-volatile data storage. Processor 155 may execute
applications or programs stored in memory 152 or secondary storage
154, or received from the Internet or other network 160. Input
device 156 may include any device for entering information into
computer 151, such as a keyboard, key pad, cursor-control device,
or touch-screen. Display device 153 may include any type of device
for presenting visual information such as, for example, a computer
monitor, flat-screen display, or display panel. Output device 158
may include any type of device for presenting a hard copy of
information, such as a printer, and other types of output devices
include speakers or any device for providing information in audio
form.
Computer 151 also includes in this example spectrometer 69
connected with fiber optic cable 68. Spectrometers are known in the
art, and the term "spectrometer" refers to any type of component
for converting a light signal into a corresponding electrical
signal at various specific wavelengths. Therefore, spectrometer 69
receives the light signal from fiber optic cable 68, converts it
into a corresponding analog electrical signal, and digitizes the
analog signal through an analog-to-digital (A/D) converter to
produce a digitized version of the raw spectral signature. The A/D
conversion can be implemented as part of spectrometer 69 or as a
separate component such as a controller card in computer 151. Also,
spectrometer 69 can be implemented as part of computer 151 or as a
separate physical component electronically linked with computer
151. Depending upon the agricultural product to be analyzed,
spectrometer 69 can be calibrated to convert a range of the light
signal between particular wavelengths. The range of the light
signal used for prediction of protein content, as in this example,
may be different for various types of agricultural products as
determined through empirical evidence.
Although computer 151 is depicted with various components, one
skilled in the art will appreciate that this computer can contain
additional or different components. In addition, although aspects
of an implementation consistent with the present invention are
described as being stored in memory, one skilled in the art will
appreciate that these aspects can also be stored on or read from
other types of computer program products or computer-readable
media, such as secondary storage devices, including hard disks,
floppy disks, or CD-ROM; a carrier wave from the Internet or other
network; or other forms of RAM or ROM. The computer-readable media
may include instructions for controlling a computer system, such as
computer 151, to perform a particular method.
Computer 151 can display a screen through which a user interacts
with the system. FIG. 12 is a diagram of a screen illustrating how
a user can interact with the system, and this screen may be
displayed on display devices associated with the user's computers.
Computer 151 can also display a screen to present information, such
as constituent data to a user, and an example of such a screen is
shown in FIG. 18. The term "screen" refers to any visual element or
combinations of visual elements for displaying information;
examples include, but are not limited to, user interfaces on a
display device or information displayed in web pages or in windows
on a display device. The screens may be formatted, for example, as
web pages in HyperText Markup Language (HTML), or in any other
suitable form for presentation on a display device depending upon
applications used to interact with the system.
The screens include various sections, as explained below, to
provide information or to receive information or commands. The term
"section" with respect to screens refers to a particular portion of
a screen, possibly including the entire screen. Sections are
selected, for example, to enter information or commands or to
retrieve information or access other screens. The selection may
occur, for example, by a using a cursor-control device to "click
on" or "double click on" the section; alternatively, sections may
be selected by entering a series of key strokes or in other ways
such as through voice commands or use of a touch screen. In
addition, although the screens shown in FIGS. 12 and 18 illustrate
a particular arrangement and number of sections, other arrangements
are possible and different numbers of sections in the screen may be
used to accomplish the same or similar functions of displaying
information and receiving information or commands. Also, the same
section may be used for performing a number of functions, such as
both displaying information and receiving a command.
The processing to support the screens is shown in the flow charts
of FIGS. 11, 13, 14, 17, and 19 specifying various routines. The
processing may be implemented in software, such as software
modules, for execution by computer 151 or other machines.
Software Processing
FIG. 1 is a flow chart of a main routine 170 for execution by
computer 70 in order to analyze constituent contents of an
agricultural product and provide a visual indication of it. The
main routine 170 shown in FIG. 11 may be used in conjunction with a
main screen 190 shown in FIG. 12. Main screen 190 may be displayed
on a corresponding display device and can be used for inputting
commands into the computer. In particular, main screen 190 includes
a section 192 for determining a constituent content in a manual
mode, a section 194 for determining a constituent content in an
automatic mode for a continuous stream of an agricultural product,
a section 195 for determining a constituent content in an automatic
mode for a stop-and-go stream of an agricultural product, a section
196 for displaying a grid map of constituent contents of a field,
and a section 198 for displaying a contour map of constituent
contents of a field.
A user may select one of these sections to enter the corresponding
command. The manual mode referred to in section 192 means that the
optical sensor analyzes and predicts a constituent content in
response to user input. The automatic (continuous) mode referred to
in section 194 means that the optical sensor, based upon a time
parameter, repeatedly analyzes and predicts a constituent content
of a continuous stream of the agricultural product. The automatic
(stop-and-go) mode referred to in section 195 means that the
optical sensor, based upon a time parameter, repeatedly analyzes
and predicts a constituent content of a stop-and-go stream of the
agricultural product. The stop-and-go stream refers to
substantially stopping the flow of the agricultural product past
the sensing window, sampling the product, and then restarting the
flow of the product.
In routine 170, the user selects a command such as is shown in main
screen 190 (step 172). Based upon the enter command, the system
executes a corresponding routine. If the user selected section 192
for the manual mode (step 174) or one of sections 194 or 195 for
the automatic modes (step 178), the system executes a determine
constituent routine (step 176). If the user selected section 196
for displaying a grid map (step 180) or section 198 for displaying
a contour map (step 184), the system executes a generate map
routine (step 182). Main screen 190 is only one example of how a
user may enter a command to the system, and a user can enter
commands in a variety of ways, such as through use of a touch
screen, voice command, key stroke, or peripheral device.
FIG. 13 is a flow chart of a determine constituent routine 200
triggered by step 176. In routine 200, the system receives a
selected mode (step 202); this mode may be determined based upon
which section 192, 194, or 195 the user selected within screen 190.
The system determines whether the user selected the manual or
automatic mode (step 204). If the user selected the manual mode,
the system waits for user input (step 218). When the system
receives an appropriate user input (step 220), it analyzes the
agricultural product to predict a constituent of it, as described
below. Ite user input can be any type of user-entered command,
either locally or remotely through network 160, to sample the
agricultural product. That command can be entered through any of
the exemplary user input devices provided above such as through a
key stroke, touching a touch-screen, or a spoken command, or
through other devices.
For the automatic mode, the system determines if the user had
selected the continuous mode identified in section 194 (step 205).
If the user selected the continuous mode, the system proceeds with
sampling the product, as described below. If the user had selected
the stop-and-go mode identified in section 195, and as determined
in step 205, then the system stops the flow of agricultural product
by sending an appropriate signal to control unit 47 controlling
motor 48 (step 207). Once the motor is stopped, the system waits
for a signal from sensor 42, indicating that the stopped flow of
the agricultural product has reached the level of the sensor (step
209), which means that the agricultural product has accumulated by
a sufficient amount to be present in the sensing window.
For both automatic modes, the system resets a timer for the time
parameter (step 206). The time parameter determines when to
repeatedly sample the agricultural product for these exemplary
embodiments. The system can alternatively repeatedly sample the
agricultural product based upon other input parameters. For
example, it can specify particular GPS coordinates to trigger
sampling. As another example, it can monitor a speed of a combine
on which sensor is affixed in order to calculate, using an internal
clock of computer151, when the combine travels a particular
physical distance and sample the agricultural product at those
distances.
The system then samples the agricultural product for any of the
manual, continuous automatic, or stop-and-go automatic modes as
follows. In this exemplary embodiment, the system obtains GPS
coordinates from GPS sensor 157 (step 208). The system then samples
the stream of agricultural product and analyzes it to predict
constituent content (step 210). This step is further explained
below. The system stores the GPS coordinates and an indication of
the corresponding constituent content (step 212). The system may
store this information in a database structure that associates GPS
coordinates with the corresponding constituent contents, such as in
a table or relational database form as shown in Table 1.
TABLE 1 GPS coordinates longitude latitude constituent content
longitude value 1 latitude value 1 constituent content value 1
longitude value 2 latitude value 2 constituent content value 2 . .
. longitude value N latitude value N constituent content value
N
The system may also display the GPS coordinates and indication of
the corresponding constituent content (step 214). This step permits
the optional display in real-time or near real-time of constituent
contents as the stream of agricultural product is analyzed. For
example, as grain is harvested in the field, a combine operator may
view an indication of protein content of the grain as it is being
harvested. As another example, as grain or other agricultural
product is moved through the assembly line of a food processing
plant, an operator may view an indication of the protein content of
the grain as it moves through the food-making process.
If the system is in the automatic mode (step 215), it also
determines whether is in the continuous or stop-and-go automatic
mode (step 217). If it is the stop-and-go mode, it restarts the
flow of the agricultural product by sending an appropriate signal
to control unit 47 controlling motor 48 in order to restart the
motor (step 219). For both automatic modes, the system determines
constituent contents at regular intervals by determining if the
timer has expired (step 216). Therefore, at regular intervals, the
system can sample the stream of agricultural product, predict the
constituent content of it, and store an indication of that
constituent content along with the GPS coordinates. If the system
is in the manual mode, it returns to step 218 to wait for another
user input.
This exemplary routine 200 is shown for use on a combine during
harvesting of a field. However, if the optical sensor were used
within a different environment, such as in a food processing
facility, certain steps may not be necessary, such as the steps to
obtain the GPS coordinates. As an alternative to obtaining the GPS
coordinates, if the stream of agricultural product is within a food
processing facility, the system could alternatively store time
indications of when the stream of agricultural product was sampled
and analyzed. The system could then display those times along with
corresponding indications of constituent content in order to
provide, for example, an indication of the consistency in the
constituent content of the agricultural product as used within the
food processing facility.
In addition, the resulting constituent content could be used as a
signal to trigger the agricultural product being directed to
different portions of a food processing facility depending upon the
type of constituent required at those places. Alternatively, the
indication of constituent content could be used for labeling of
different types of food products produced from the agricultural
product. For example, a stream of grain having a high protein
content may be directed to or labeled as a high quality food
product, while grain having a lower protein content could be
directed to or labeled as a lower quality food product.
FIG. 14 is a flow chart of a routine or algorithm 230 for
performing the analysis of the sampled agricultural product in step
210. This algorithm is a technique that allows processing of the
acquired signal for predicting the constituent of the product. This
technique does not require a reflectance or transmission signal of
a separate standard, to process the acquired signal, as being done
by standard methods. This algorithm can work with both reflectance
and transmission spectral signals. This algorithm is called
prediction calibration using product spectrum.
In routine 230, the optical sensor obtains the raw spectral and
dark signals (step 232). The raw spectral signal is obtained from
the transmitted light through the agricultural product, as received
by the sensor and directed onto fiber optical cable 68. The
received signal is in analog form at that point.
FIGS. 15 and 16 are examples of a raw spectral signal 242 (with the
dark signal subtracted) and various points upon it corresponding to
particular wave lengths. The raw spectral signal may be analyzed
between wavelengths corresponding to points 244 and 250. These
wavelengths may be known in advance through various techniques to
determine a spectral range required for predicting constituent
content of a particular agricultural product.
A point 248 represents a particular wavelength used as a reference
to normalize the data points. In particular, it may be determined
based upon empirical evidence that a particular constituent is not
affected by the magnitude of the waveform at a particular
wavelength.
Referring again to FIG. 14, the analog signal is digitized to
produce a digitized raw spectral signals (step 234). To perform
processing on the digitized signal, the computer can store a
particular number of the data points from the digitized signal
within a table, array, or other data structure corresponding with
magnitudes of the signal at certain wavelengths. In addition, the
computer can store a dark signal, typically predetermined, for the
signal at the corresponding wavelengths. Use of a "dark signal" is
known in the art. It can also store a resulting (spectra-dark)
signal as explained below. Table 2 illustrates an exemplary array
of data points.
TABLE 2 magnitude of magnitude of spectra - wavelength received
signal dark signal dark signal .lambda..sub.k magnitude value 1
dark signal magnitude value 1 value 1 .lambda..sub.k+1 magnitude
value 2 dark signal magnitude value 2 value 2 . . . .lambda..sub.L
magnitude value N dark signal magnitude value N value N
Next, the dark signal is subtracted from the raw spectral signal
(step 236) using, for example, equation (1). ##EQU1##
The computer can perform the subtraction for each data point by
retrieving each value of the digitized spectral signal and dark
signal from a data structure such as shown in Table 2 and
sequentially performing the subtraction. It can store the resulting
values back in column four of Table 2 as shown for use in
subsequent processing, or in another data structure.
Alternatively for step 236, the computer can determine an average
signal for each data point. In particular, the system can take two
readings of the same sample and determine the average of the
sampled signal minus the dark signal. For each data point
(spectra-dark) shown in equation (1), the computer determines
Average_signal=[((signal1-dark)+(signal2-dark))/2]. Signal1 and
signal2 represent the two raw spectral signals of the same sample
for each data point. The system can also take more than two
readings for each sample and determine the average based upon more
than two readings. The use of an average signal for the same sample
(each data point) helps to remove variations in the signal
resulting from random particle size; other mathematical techniques
can also be used to help remove or minimize effects of those
variations.
Next, the (spectral-dark) signal is normalized using a reference
wavelength to produce a normalized signal (step 238). This step can
be accomplished using, for example, equation (2). ##EQU2##
where, Sn.sub..lambda. =normalized signal from .lambda.=k to
.lambda.=L .lambda..sub.d =normalizing wavelength .lambda..sub.d
S.sub..lambda.d =spectral signal at normalizing wavelength
.lambda..sub.d .lambda..sub.d can be .lambda..sub.1, or
.lambda..sub.2, or .lambda..sub.3, . . . .lambda..sub.n, pr
.lambda..sub.T S.sub..lambda.d can also be P where,
P=f(.lambda..sub.1, .lambda..sub.2, .lambda..sub.3, . . .
.lambda..sub.n) f=any mathematical function
As shown in FIG. 15, .lambda..sub.1, .lambda..sub.2, . . .
.lambda..sub.n (339, 340, 341) are wavelength(s) critical for
predicting the constituent of the product. They can be
wavelength(s) with highest or higher correlation with the
concentration of the constituent, to be determined. As also shown
in FIG. 15, .lambda..sub.T is a reference wavelength or equivalent
of a band of wavelengths that does not contribute to the
concentration of the desired constituent or does not have any
correlations with the concentration of the desired constituent. The
wavelength .lambda..sub.T could be predetermined from prior
experiments, or through empirical evidence.
In equation (2), S.sub..lambda.d can be replaced by S.sub..lambda.b
as shown in equation (3). ##EQU3##
f=any linear and/or non-linear function, that may or may not be
dependent on .lambda., wavelength.
A specific form of equation (2) can be obtained by using a
.lambda..sub.T for .lambda..sub.d and a specific function (average)
in equation (3), as shown in equation 3-A. ##EQU4##
where N=number of observations corresponding to wavelength,
.lambda.=q, . . . t.
FIG. 16 illustrates this concept further. It can be seen that,
instead of using a reference wavelength, a band of wavelengths
centered around the reference wavelength may be used. In
particular, a spectral signal 252 may have a reference wavelength
at point 256. In order to account for minor variations in the
intensity at point 256 based on other factors apart from
constituent content, the system may use an average magnitude value
of a band of wavelengths between points 254 and 258 to normalize
the spectral signal.
With this routine, the reference wavelength is known in advance.
The magnitude at the reference wavelength can thus be used to
normalize the spectral signal and predict constituents in real-time
or near real-time without the need to obtain separate reference
signals upon every sampling of the agricultural product.
By using a data structure, the computer can retrieve the values of
the data points for each wavelength and compute the processed
signal (S.sub.n.lambda.). It can also store the resulting processed
signal values in the same table or other data structure, as
illustrated by Table 3.
TABLE 3 wavelength processed signal .lambda..sub.k value 1
.lambda..sub.2 value 2 . . . .lambda..sub.L value N
The system then processes the normalized signal using, for example,
a standard prediction technique known in the art to determine the
constituent content (step 240), as explained below.
Additional processing can be performed on the data points as well.
In addition to the processing occurring through equation (1), the
(spectra-dark) signal can be processed in any linear or non-linear
way before the normalizing step 238. Also, the normalized signal
can be processed in any linear or non-linear way in addition the
normalizing occurring through equation (2).
FIG. 17 shows a flow chart for two other routines or algorithms 260
that will also eliminate the need for obtaining reflectance or
transmittance signals of separate standards and thus are alternate
embodiments for analyzing the spectral signal to determine the
constituents of the agricultural product in step 210. In both
techniques (alternate embodiments), the system first obtains a
reference signal and dark signal (step 262). To obtain the
reference signal, a separate physical standard is not required.
These two techniques are called (1) prediction calibration using
specific wavelength and (2) prediction calibration using
illumination spectrum.
The technique "prediction calibration using specific wavelength"
involves receiving the radiation transmitted without any
agricultural product present in the optical sensing window and
using the magnitude of that received radiation at specific
wavelengths. The other technique, "prediction calibration using
illumination spectrum," involves receiving transmitted radiation
with a gating mechanism or a mechanism to reduce light intensity
while no agricultural product is present in the optical sensing
window and using the magnitude of that received radiation. A gating
mechanism could include any device to block or attenuate at least a
portion of the transmitted radiation and examples include a mesh
screen. The mechanism to reduce light intensity could also include
any method to decrease the intensity of the illumination source
(light) by changing the supplied voltage or current. In this
technique, a gating mechanism or mechanism to reduce light
intensity is only used to obtain the reference signal (step 262).
For obtaining the spectral signal of the product sample (step 264),
a gating mechanism or mechanism to reduce light intensity is not
used.
For both of these techniques, the apparatus can include a valve or
other mechanism to block the stream of the agricultural product
passed through the optical sensing window in order to obtain the
reference signal.
The system can obtain the reference signal through these or other
methods at various times. It can obtain the reference signal upon
each irradiation and analysis of the agricultural product as shown
in routine 260. Alternatively, it can obtain the signal other than
during every analysis. For example, it may obtain the reference
signal once for the analysis of an entire field of the agricultural
product during harvesting. It may also obtain the reference signal
periodically based upon, for example, a user-defined time
parameter.
The system then performs essentially the same steps as in routine
230 except that it is uses the reference signal to normalize the
spectral signal. In particular, the system obtains the raw spectral
signal through the stream of agricultural product (step 264),
digitizes it to produce a digital signal having a series of data
points corresponding to particular wavelengths (step 266),
subtracts the dark signal from the raw spectral signal (step 268),
and normalizes the (spectral-dark) signal using the reference
signal in order to produce a normalized signal (step 270). The
processing for steps 268 and 270 depend upon a particular
application and the type of reference signal used.
For the prediction calibration using specific wavelength method,
equation (4) can be used to perform steps 268 and 270. ##EQU5##
In equation (4).lambda..sub.1, .lambda..sub.2 . . . are wavelengths
critical for predicting the constituent of the agricultural
product,
I.sub.t.lambda.2 is the magnitude of the transmitted radiation (raw
spectral signal) at wavelength .lambda..sub.2 with the agricultural
product present in the sampling window,
I.sub.t.lambda.1 is the magnitude of transmitted radiation (raw
spectral signal) at wavelength .lambda..sub.1 with the agricultural
product present in the sampling window,
I.sub.o.lambda.2 is the magnitude of the received radiation at
reference wavelength .lambda..sub.2 (reference signal) with no
agricultural product present in the sampling window,
I.sub.o.lambda.1 is the magnitude of the received radiation at
reference wavelength .lambda..sub.1 (reference signal) with no
agricultural product present in the sampling window.
For the prediction calibration using illumination spectrum method,
equation (5) can be used to perform steps 268 and 270. ##EQU6##
In equation (5), ##EQU7##
is the reference signal obtained for wavelengths k to L with the
gating mechanism but no agricultural product present, and
##EQU8##
is the transmitted raw spectral signal obtained for wavelengths k
to L without the gating mechanism but with the agricultural product
present.
For step 268, the computer can alternatively determine an average
signal for each data point and use that average in the equations as
described above. In particular, the system can take two readings of
the same sample and determine the average of the sampled signal
minus the dark signal. For each data point, shown in equation (4)
or (5), the computer determines
Average_signal=[((signal1-dark)+(signal2-dark))/2]. Signal1 and
signal2 represent the two raw spectral signals of the same sample
for each data point. The system can also take more than two
readings for each sample and determine the average based upon more
than two readings. The use of average signal for each data point
helps to remove variations in the signal resulting from random
particle size; other mathematical techniques can also be used to
help remove or minimize effects of those variations.
The system processes the normalized signal using, for example, a
standard prediction technique known in the art to determine the
constituent content (step 272).
The standard techniques for steps 240 and 272 usually involve
linearizing the data, which can be accomplishing by calculating a
logarithm of the data output from step 238 for routine 230 or step
270 for the alternate embodiments of routine 260. The logarithm
involves calculating S.sub.L =log.sub.10 (S.sub.n.lambda.) for ale
each of the data points. The signal S.sub.c (equation (4)) was
already linearized. The set of linearized data points S.sub.L or
S.sub.c can then be processed according to standard techniques for
the particular constituent desired and type of agricultural product
sampled. For example, to calculate the protein content of barley,
oats, rye, triticale, and wheat of all classes, the data points
S.sub.L or S.sub.c can be processed according to the American
Association of Cereal Chemists (AACC) method 39-10, and to
calculate the protein content of whole-grain wheat, the data points
S.sub.L or S.sub.c can be processed according to AACC method 39-25.
AACC methods 39-10 and 39-25 are known in the art. These standard
techniques can involve statistical analysis of the data points,
which can be accomplished using conventional software for
performing the particular statistical analysis functions on the
data.
For other types of constituents or agricultural products, other
standard techniques can be used to process the data points S.sub.L
or S.sub.c. Accordingly, once of the data points are obtained after
step 238 in routine 230 and step 270 in routine 260, those data
points can be processed in steps 240 and 272 according to a variety
of standard and known techniques for a variety of constituents and
agricultural products. The data points can also be stored for later
processing or for processing according to varying standards and
statistical techniques.
As an example, the S.sub.L data points were tested and processed as
follows for determining protein content of wheat. Two prediction
calibration techniques were used. They were (1) prediction
calibration using product spectrum and (2) prediction calibration
using illumination spectrum. This experiment also was used to
validate the performance of these two prediction calibration
techniques or algorithms. Using GRAM-32, version 5.03, a software
program from Galactic Industries, N.H., the signal S.sub.L was
further processed as follows: offset correction was performed on
the y-axis with the option of "set point to zero"; the signal was
smoothed using "binomial smoothing" with the number of points equal
to four; the data was reduced by a factor of six-to-one using an
averaging technique under the interpolation option of the software;
and a second derivative was calculated with a gap value equal to
twenty.
Randomly selected wheat samples (501 samples) were used and their
raw signals were obtained using the sensor system. After the
processing of S.sub.L, 125 data points represented each spectral
signal of every wheat sample. For the prediction calibration using
product spectrum technique, there were 420 training and 81 test
data sets. For the prediction calibration using illumination
spectrum technique, there were 420 training and 80 test data sets.
For purposes of the testing, signals were acquired in the
stop-and-go mode. Each sample was put into the sensing window two
times and two spectral signals were acquired for each sample; the
average of the two signals was used.
The data were then further processed using Principle Component
Regression (PCR) and Partial Least Square (PLS) to develop a
prediction model, using SAS software from Statistical Analysis
Software, N.C. The prediction model predicted the protein content
percentage for the wheat sample. Though the actual protein contents
of wheat samples are determined by using standard reference methods
such as the Kjeldahl (AACC 39-25) method, in this study a separate
commercial laboratory scale NIR analyzer (model Infratec 1226 grain
analyzer, FOSS, Eden Prairie, MN) was used to determine the protein
content of the wheat sample. Because this instrument is also widely
used for determining protein content, the obtained protein contents
were considered as actual protein contents of wheat samples. The
protein contents determined by the commercial NIR instruments were
compared with the protein contents predicted by the system and the
models. This testing illustrates only one example of processing the
data points, and other conventional techniques and software
programs can be used.
FIG. 18 is a diagram of a screen 280 for displaying in real-time or
near real-time GPS coordinates and corresponding constituent
content, in this example protein, in step 214 in routine 200. In
particular, a screen 280 may be displayed on a display device and
presented to an operator during analysis of the stream of
agricultural product. It includes a section 282 for providing an
indication of protein content such as a percentage or other
numerical indication. It also includes sections 284 and 286 for
displaying, respectively, longitude and latitude coordinates
providing an indication of the approximate geographic location of
the agricultural product corresponding to the displayed protein
content. Therefore, the screen can be displayed in the cab of a
combine during harvesting of an agricultural product. It can also
be displayed on a display device at a location in a food processing
facility where the stream of grain is analyzed, although it could
display a time indication rather than GPS coordinates for
implementation in a food processing facility.
FIG. 19 is a flow chart of a generated map routine 290 for
generating a visual map of constituent content among a field of an
agricultural product. Routine 290 may be triggered by user's
selection of sections 196 or 198 in main screen 190 for generating
and displaying a map. In routine 290, the system determines the
type of requested map (step 292). This may be accomplished by
determining which section the user selected in main screen 190 for
obtaining a map.
FIGS. 20 and 21 illustrate, respectively, examples of a grid map
and a contour map for providing a visual indication of protein
content or other constituent among a field of an agricultural
product. In a grid map 320 in FIG. 20, the map includes a first
axis 322 for displaying an indication of longitude coordinates, and
a second axis 324 for displaying an indication of latitude
coordinates. The latitude and longitude coordinates are divided
into a plurality of cells, such as cell 326, and an indication of
protein content is displayed for each cell. The indication of
protein content may be a number indicating a percentage of protein
content within the agricultural product for that section of the
field, or could be some other type of indication.
FIG. 21 illustrates a contour map 330 providing another way of
indicating protein content of a field. Contour map 330 includes a
first axis 332 providing longitude coordinates, and a second axis
334 providing latitude coordinates. Among those longitude and
latitude coordinates various indications are provided for different
protein content within the field. For example, a section 338
provides an indication of one level of protein content. The contour
map may be associated with a key 336 indicating which visual
indications correspond with which protein content. Although various
types of cross-hatching are illustrated for indicating protein
content, other types of visual indication may be provided, such as
different colors or shades.
Referring back to FIG. 19, grid map 320 and contour may 330 are
generated in routine 290 as follows. For the contour map, as
determined in step 308, the system retrieves the GPS coordinates
and corresponding constituent contents from a database in memory
such as is shown in Table 1 (step 310). The system extrapolates the
data to generate lines dividing constituent areas within a map of
the field (step 312). Data extrapolation techniques are known in
the art. Each area is filled in with an indication of constituent
content, such as using different colors, cross-hatching or shadings
(step 314). The system then displays the map with the GPS
coordinates and constituent content of each area (step 316), and
the system may also store the map in order to avoid having to
repeatedly generate it.
If the user had selected the grid map, as determined in step 294,
the system divides the map into cells (step 296). Each cell may be
indicated by the corresponding GPS coordinates within that cell.
The system retrieves GPS coordinates and constituent contents for a
cell from a database in memory such as is shown in Table 1 (step
298). The average constituent content is then determined for the
cell (step 300). The average may be determined by adding all the
constituent contents within that cell and dividing by the total
number of samples for that cell. The system stores an indication of
the cell and corresponding constituent content (step 302). It then
determines if more cells exist to process (step 304). If so, it
repeats step 298, 300 and 302 to determine an average constituent
content for the next cell. When all cells have been processed, the
system displays the grid map with the GPS coordinates and average
constituent content of each cell (step 306), and it may also store
the grid map to avoid having to repeatedly generate it.
The grid map and contour map may be displayed on a display device
associated with computer 70 or printed out in bard copy form.
While the present invention has been described in connection with
an exemplary embodiment, it will be understood that many
modifications will be readily apparent to those skilled in the art,
and this application is intended to cover any adaptations or
variations thereof. For example, different types of materials for
the device, and various types of software algorithms for processing
the signal resulting from irradiation of the agricultural product,
may be used without departing from the scope of the invention. This
invention should be limited only by the claims and equivalents
thereof.
* * * * *